[Wannier] Cu2ZnSnSe and Cu2ZnSnS (Boltzwann Calculations)

Wed Oct 19 14:38:45 CEST 2016

Thanks for the reply,

I gave the value at zero chemical potential and 300 K, apologies for not being clear about that.

Initially I checked for convergence by comparing the bandstructures from the Wannier interpolation and the VASP SCF calculation. It seems it is a k-point issue, as I have increased the k-mesh and do achieve a much better agreement (A valuable lesson for me).

Your help and suggestions are much appreciated.

With Regards

Alex Aziz

________________________________
From: Wannier <wannier-bounces at quantum-espresso.org> on behalf of Giovanni Pizzi <giovanni.pizzi at epfl.ch>
Sent: 19 October 2016 12:46
To: wannier at quantum-espresso.org
Subject: Re: [Wannier] Cu2ZnSnSe and Cu2ZnSnS (Boltzwann Calculations)

Dear Alex
if the band structure is the same, the results should be comparable.

There are two possible sources of difference that come to my mind:
1. The Wannier-interpolated band structure is different (e.g. because the Wannierization is not converged). Did you check if WFs are real, and the interpolated band structure is the same as the one calculated with VASP?

2. The interpolation procedure is different in the two codes. Did you check in both codes that the results do not depend on the density of k-points used for the interpolation/integration?

Also, the value that you state is evaluated at which chemical potential and temperature?

Giovanni

On 10/17/2016 03:27 PM, Alex Gal Aziz wrote:

Thanks for the message,

I have done a check with one calculation with BoltzTRAP and Boltzwann using PBE+U. Although the trend seems to be the same, the values are not really the same. The BoltzTRAP calculation gives values about 100 μV/K higher. Around 500 μV/K for BoltzTRAP compared to 400 μV/K for Boltzwann.

I believe the error must be on my part, but changing the number of WF doesn't change the result. I have kept the Bandstructure frozen till about 1 eV above the lowest conduction band.

Any ideas would be welcome.

Thanks

Alex Aziz

PhD Student

University of Reading

________________________________
From: Wannier <wannier-bounces at quantum-espresso.org><mailto:wannier-bounces at quantum-espresso.org> on behalf of Nicki Frank Hinsche <nicki.hinsche at physik.uni-halle.de><mailto:nicki.hinsche at physik.uni-halle.de>
Sent: 01 September 2016 21:38
To: wannier at quantum-espresso.org<mailto:wannier at quantum-espresso.org>
Subject: [Wannier] Cu2ZnSnSe and Cu2ZnSnS (Boltzwann Calculations)

Hej Ali,

this depends heavily on the charge carrier concentration. The difference due to bipolar effects will be way more pronounced at high temperatures and/or low charge carrier concentrations. The size of the gap merely plays a role in the degenerate limit.

bests
Nicki

--------------------------------------------------------
Nicki Frank Hinsche, Dr. rer. nat.
HC Ørsted Postdoc
CAMd, Technical University of Denmark
Fysikvej, Bld. 307, Room 208
2800 Kgs. Lyngby, Denmark
Tel.: ++45 45 25 31 84
--------------------------------------------------------

> Message: 1
> Date: Thu, 1 Sep 2016 10:03:04 +0000
> From: Alex Gal Aziz <A.G.Aziz at pgr.reading.ac.uk><mailto:A.G.Aziz at pgr.reading.ac.uk>
> To: "wannier at quantum-espresso.org"<mailto:wannier at quantum-espresso.org> <wannier at quantum-espresso.org><mailto:wannier at quantum-espresso.org>
> Subject: [Wannier] Cu2ZnSnSe and Cu2ZnSnS (Boltzwann Calculations)
> Message-ID:
>        <VI1PR0101MB2365665CF6E0025C9742F44AD1E20 at VI1PR0101MB2365.eurprd01.prod.exchangelabs.com><mailto:VI1PR0101MB2365665CF6E0025C9742F44AD1E20 at VI1PR0101MB2365.eurprd01.prod.exchangelabs.com>
>
> Content-Type: text/plain; charset="iso-8859-1"
>
> I am investigating Cu2ZnSnSe4 and Cu2ZnSnS4.
>
>
> I have been using VASP compiled with Wannier functions to generate the initial projections and then generate the MLWF using the Disentanglement procedure. I have checked the band structures and they look correct with the appropriate band gaps (~1.5 for Cu2ZnSnS4 and 0.9 Cu2ZnSnSe4).
>
>
> When I check the Seebeck using BoltzWann I seem to get similar results in both cases. I am trying to figure out if this seems plausible due to the difference in band gaps.
>
>
> Any suggestions would be welcomed
>
>
> Thanks
>
> Alex Aziz
>
>
> PhD Student
>
> Department of Chemistry, University of Reading
> Whiteknights, Reading RG6 6AD

_______________________________________________
Wannier mailing list
Wannier at quantum-espresso.org<mailto:Wannier at quantum-espresso.org>
http://mailman.qe-forge.org/cgi-bin/mailman/listinfo/wannier
Wannier Info Page - mailman.qe-forge.org<http://mailman.qe-forge.org/cgi-bin/mailman/listinfo/wannier>
mailman.qe-forge.org
To see the collection of prior postings to the list, visit the Wannier Archives. Using Wannier: To post a message to all the list members, send email ...

_______________________________________________
Wannier mailing list
Wannier at quantum-espresso.org<mailto:Wannier at quantum-espresso.org>
http://mailman.qe-forge.org/cgi-bin/mailman/listinfo/wannier

--
Giovanni Pizzi
EPFL STI IMX THEOS
MED2 1326
Station 9
CH-1015 Lausanne (Switzerland)
Phone: +41 21 69 31124
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20161019/74409a0c/attachment.html>